View
213
Download
0
Category
Preview:
Citation preview
City, University of London Institutional Repository
Citation: Govender, R., Lee, M. T., Drinnan, M., Twinn, C., Davies, T. & Hilari, K. (2016). Psychometric Evaluation of the Swallowing Outcomes after Laryngectomy (SOAL) Patient-reported Questionnaire. Head and Neck, 38(S1), E1639-E1645. doi: 10.1002/hed.24291
This is the accepted version of the paper.
This version of the publication may differ from the final published version.
Permanent repository link: http://openaccess.city.ac.uk/12471/
Link to published version: http://dx.doi.org/10.1002/hed.24291
Copyright and reuse: City Research Online aims to make research outputs of City, University of London available to a wider audience. Copyright and Moral Rights remain with the author(s) and/or copyright holders. URLs from City Research Online may be freely distributed and linked to.
City Research Online: http://openaccess.city.ac.uk/ publications@city.ac.uk
City Research Online
1
PSYCHOMETRIC EVALUATION OF THE SWALLOWING OUTCOMES AFTER LARYNGECTOMY (SOAL) PATIENT-REPORTED OUTCOME MEASURE
Abstract
Objectives: To evaluate the psychometric properties of the Swallowing Outcomes After Laryngectomy (SOAL)
patient-reported outcome measure in a large group of people with laryngectomy.
Design: cross-sectional psychometric study.
Participants: Laryngectomy patients (minimum 3-months post-treatment) attending routine hospital follow-
up.
Main outcome measure: psychometric evaluation of SOAL.
Results: One hundred and ten people participated. Thirteen percent had a laryngectomy, 63% had
laryngectomy with radiotherapy, and 24% had laryngectomy with chemoradiation therapy. The SOAL showed
good quality of data (minimal missing data and floor effects); good internal consistency (=.91); and
adequate test-retest reliability (intra-class correlation coefficient =.73). In terms of validity, it differentiated
people by treatment group (F(2,85)=8.02, p=0.001) and diet texture group (t(102)=-7.33, p<0.001).
Conclusions: The SOAL demonstrates good validity and has potential for use in research. Further study is
required to determine its clinical application.
2
INTRODUCTION
Measures of patient-reported functional outcomes have an established place in clinical research, but have
more recently also been explored for their use in routine clinical practice. 1,2 In the field of head and neck
cancer, patient reported outcome measures can be useful tools in prompting discussions between clinician
and patient about rehabilitation priorities and clinical interventions so that the greatest concerns for patients
are given due attention during their follow-up visits. 3,4 Such tools must demonstrate good reliability and
validity if they are to be widely used in clinical practice and/or research. 5,6
There have been a number of tools developed, focusing on dysphagia-related symptoms and dysphagia-
related quality of life 7-10. All of these questionnaires were developed for a more general head and neck
population and all assume the presence of a larynx. They highlight aspects of swallow that are not always
relevant in the absence of a larynx and may not always capture swallowing changes post total laryngectomy.
Currently, there are no validated tools specifically addressing swallowing outcomes after total laryngectomy.
The Swallowing Outcomes After Laryngectomy (SOAL) was developed as the first laryngectomy specific
measure to report swallowing problems experienced by this subset of patients with head and neck cancer
who have their larynx surgically removed. 11 In addition to the loss of laryngeal voicing, laryngectomees also
have altered swallowing and respiratory anatomy and physiology. 12 A questionnaire that takes account of
these unique changes was devised in consultation with a patient focus group and expert clinicians.
The SOAL is a 17-item scale listing problems people may experience with their swallowing after
laryngectomy. Full details of the questionnaire development and preliminary validation in a small sample of
patients are described in an earlier paper.11 Notably, only 19 of the 58 patients in the preliminary validation
had undergone a total laryngectomy with the remaining patients representing the dysphagic and non-
dysphagic groups. The SOAL was shown to have good discrimination for known groups testing. Dysphagic,
non-dysphagic (non-complaining volunteers) and laryngectomy groups demonstrated significantly different
scores (p<0.001). The non-complaining group showed very low scores (least impaired) and the dysphagic
group showed much higher scores (most impaired). The laryngectomy group demonstrated a range of scores,
which correlated well with the type of diet recorded: patients with the lowest SOAL scores were eating a
normal diet whilst those having a very soft/liquid diet and supplements had higher scores. A relationship
between the SOAL and an instrumental measure of swallowing was also demonstrated by a positive and
significant correlation (r=0.5; p=0.03) between SOAL and a modified barium swallow checklist score.11
The findings from our earlier work was based on a small sample of laryngectomees. The current study
therefore aims to evaluate further the psychometric properties of the SOAL in a larger sample. In particular,
3
we have evaluated the quality of the data collected on the measure, its reliability (internal consistency and
test-retest), and construct validity in a generic sample of laryngectomy patients under hospital follow-up.
METHODS
Participants & procedures
We carried out a cross-sectional, questionnaire based, psychometric study. Ethical approval was obtained
from an NHS multicentre Research Ethics Committee. Recruitment took place over an 18-month period in
four NHS hospitals. Patients over 18 years of age who had undergone a total laryngectomy, were a minimum
of 3 months post their last oncological treatment, and who had no known head and neck recurrent disease
were eligible to take part. Individuals with extended laryngectomy (eg flap reconstruction) were also
included. Ability to understand English was necessary for participation. Patients were excluded if they had: a
partial laryngectomy, which did not include a complete separation of the trachea and oesophagus and
creation of a stoma; other known conditions which affected swallowing (eg neurological disease); or if they
were unable to provide informed consent.
Participants were recruited by speech and language therapy clinicians, during follow-up clinics. All clinicians
who contributed to data collection in the study were members of a Head and Neck Special Interest Group.
This forum afforded the opportunity for a group training session to ensure a level of consistency in the data
collected across the hospital sites. Clinicians were fully apprised of the inclusion/exclusion criteria, and the
demographic and treatment data to be obtained from the medical notes. Written information following the
training session was sent to all clinicians for reference at any stage in the study. A system was put in place for
queries during recruitment to be directed to the chief investigator so that any relevant information could be
cascaded to all sites.
Recruitment occurred face to face during routine follow-up clinics. The protracted period of recruitment
allowed clinicians to ascertain when laryngectomy patients were scheduled for their clinic visit and to plan
ahead thereby maximising recruitment in this minority population. Following a brief explanation about the
study, a patient information leaflet and a consent form were given to the patient usually whilst still in the
waiting room. Patients were given the option of taking the information away and returning the questionnaire
by post, or consenting and completing the questionnaire while they waited. Patients who agreed to
participate had the choice of completing the questionnaire on their own or having the clinician go through
the questionnaire with them. This allowed for inclusion of those patients who had difficulty reading or simply
preferred an interview style to self-completion of the questionnaire. Clinicians were instructed to read the
questions verbatim to minimise any differences that could occur between participants who chose to
complete the questionnaire independently and those who required the clinician to read out the questions.
4
Clinicians were advised to clearly explain the response format at the outset and only to provide clarification
when requested. It was also explained to the patient that any discussion about their swallowing could be
expanded upon after completion of the questionnaire. Based on our preliminary paper, completion of the
SOAL requires 5- 15 minutes 11. Patients returned their questionnaires on the same day. In cases where time
prohibited questionnaire return on the same day or if patients chose to take the questionnaire away, a
stamped addressed envelope was provided for the patient to return the questionnaire by post. Demographic
details and treatment information was obtained by the clinician from the medical notes. Test-retest reliability
data were collected from one hospital site with a 2-week test-retest interval time. Participants invited to
complete a repeat questionnaire were chosen on the basis that they were known to be clinically stable. This
decision was based on no recent hospital admissions and no reports of new symptoms noted in the medical
notes over the previous 2 months.
The Swallowing Outcome After Laryngectomy Questionnaire
A copy of the questionnaire is attached (Appendix 1). It consists of 17-items presented on a single page. It
has a 3-point response scale (0 = No, 1 = A little, 2 = A lot). Whilst this is potentially a limited range, the
format derived from patient descriptors that emerged during the initial focus groups was retained in the final
questionnaire. Scores range from 0-34 with higher scores reflecting greater self reported problems. The
direction of this scoring system is consistent with other symptom burden questionnaires such as the Sydney
Swallow Questionnaire 8,9, where higher scores reflect greater symptom burden or poorer swallow function.
Psychometric evaluation and data analysis
All completed questionnaires were returned to a central site. Responses were transferred to an electronic
template and populated into an excel spreadsheet. Data input was done by two speech and language
therapists to improve accuracy. Additionally, a third clinician performed random checks on 10% of the data
entries to observe for any errors. Data analysis was performed using SPSS v19.
Standard psychometric methods 13,14 were used to evaluate the quality of the data, internal consistency, test-
retest reliability and construct validity (known groups), using a previously developed framework 15,16. The
criteria adopted for this study and summarised below is based on the framework outlined by Lamping et.al.
(2002) 15
Quality of data, internal consistency and test-retest reliability
Quality of data is evaluated by the completeness of the data and the distribution of scores: missing data
should be <10%; floor effects should be <80%, and ceiling effects should be <80% (ie, frequency of
5
endorsement / percentage of people choosing the bottom and top end of the response scale); and skewness
values should range between 1 and -1 (meaning data is normally distributed) for 75% of questionnaire items
(some skewness is expected post-laryngectomy). Internal consistency reflects the homogeneity of the scale,
i.e. all items measuring the same underlying construct: criteria were Cronbach’s alpha >.70 and item total
correlations ≥.30. Test-retest reliability is about the stability of the measure when administered twice across
time, when no change is expected: total score intra-class correlation coefficient (ICC) >.75.
Construct validity
Factor structure: within-scale analyses should show that a single entity is measured and that items can be
combined to give an overall score. As well as internal consistency and high item-total correlations, evidence
from principal component factor analysis was sought to demonstrate that a single construct is being
measured: in unrotated Principal Components Analysis (PCA) items should load >.30 on the first component.
Additionally, factor analysis (Principal Axis Factoring, PAF) was undertaken to explore whether there was an
underlying factor model (whether items grouped into subdomains). A sound factor model should be
conceptually clear and meet the following criteria: 17,18 Items should load ≥0.40 and should not cross-load
(i.e., load on ≥2 factors with values ≥0.4 and with a difference of <0.2 between them) and there should be at
least 3 items per factor.
Known groups validity: We evaluated known group differences by testing two hypotheses: SOAL scores will
be better for people who have undergone simple laryngectomy than for those who also had radiotherapy or
chemo-radiation therapy. This was analysed with an independent groups ANOVA followed by pairwise
comparison with Tukey correction. Additionally, SOAL scores will be better for those having a normal diet
than those on a modified diet or no oral intake. An independent samples t-test was used to compare the
normal diet group vs the combined modified diet and no oral diet groups.
RESULTS
Participants
Questionnaire responses were obtained from 110 participants across the 4 hospital sites. Three patients who
were eligible and approached for participation declined. We did not systematically collect information on
which patients required assistance to complete the questionnaire. The majority of patients returned the
questionnaire on the same day. Of the small number that took the questionnaire away, only one patient
failed to return it via the post following a telephone reminder. This provided a response rate of 96.5%. Table
one presents the participant characteristics.
[Table 1 about here]
6
20 patients were invited to participate in the test-retest reliability subsample; 19 (95%) returned both
questionnaires. As total SOAL scores were used in the analysis, incomplete questionnaires had to be omitted.
4 patients were therefore excluded due to missing data. 15 patients (11 males and 4 females) who ranged in
age from 57 to 71 with a mean (SD) = 65.2 (4.04 ) were included in the test-retest analysis.
Psychometric properties
Quality of data, internal consistency and test-retest reliability (see table two)
In terms of quality of data, three items showed floor effects, meaning that few patients reported these
difficulties (problems swallowing thin and thick liquids and liquids sticking in throat). No items showed
ceiling effects. Five item distributions (29%) were positively skewed (problems swallowing thin and thick
liquids; liquids sticking in the throat; problems eating soft food; problems eating due to dry mouth). There
were no floor or ceiling effects and no skewness in the overall SOAL scores. Although no items failed the
criterion for missing data, for five of 110 participants (4.5%) we could not calculate an overall SOAL score,
because of missing data.
The SOAL showed high internal consistency (Cronbach’s alpha = 0.91). Item-total correlations ranged from
0.38 to 0.77. Test-retest reliability was acceptable (ICC = .73).
[table 2 about here]
Validity
Factor structure: On PCA, all items loaded with values > 0.44 on the first component, confirming that a single
construct was being measured. In PAF factor analysis three factors had Eigen values >1 (50.6% of variance
explained) but the scree plot indicated a two-factor structure (46% of variance explained). Inspecting the
factors did not reveal a conceptually clear and statistically robust subdomain structure.
Known groups validity (see table 3): Results confirmed our hypothesis of different SOAL scores for different
treatment groups. Mean (SD) SOAL scores were significantly different (F(2,85)=8.02, p=0.001) with those
with a simple laryngectomy having the best scores [7.4 (7.7)] followed by those with additional radiotherapy
[10.2 (6.2)] and those with additional chemoradiation therapy [16.6 (8.8)]. The effect size was large (η p2 =
0.16) suggesting a large difference between the groups. Pairwise comparisons with Tukey correction showed
there was no significant difference between the simple laryngectomy group and the laryngectomy with
additional radiotherapy group (p=0.44). However, those treated with laryngectomy and additional
7
chemoradiation therapy had significantly worse SOAL scores than those with simple laryngectomy (p=0.002)
and those with laryngectomy and radiotherapy (p=0.002). Additionally those on a normal textured diet had
significantly better SOAL scores [mean (SD) = 8.5 (6.0)] than those on modified/no oral intake [mean (SD) =
18.3 (6.5)] (t(102)=-7.33, p<0.001).The effect size was large (d=-1.55) suggesting a substantial difference
between the two groups.
[table 3 about here]
Discussion
In this study of 110 laryngectomy patients, we examined further the psychometric properties of the SOAL
that was initially tested in a smaller sample of patients). 11 The SOAL showed good quality of data (minimal
missing data and floor effects); good internal consistency and acceptable test-retest reliability. It
differentiated patients by treatment group and diet group demonstrating good Known groups validity. No
items on the questionnaire failed the criterion for missing data which suggests that respondents understood
and were able to complete all questions. This was consistent with our findings in the preliminary validation
when face validity was established during the development phase of SOAL.
The three items that demonstrated floor effects (problems swallowing thin and thick liquids, liquids sticking
in the throat) were predictable as most laryngectomees are able to swallow liquids relatively well. It is
therefore unsurprising that more than 80% of our sample of follow-up patients reported no problems with
swallowing liquids. Problems eating soft foods and problems eating due to dry mouth also demonstrated
positive skewness. This too might be expected given that the target radiation fields in treating laryngeal
cancer will most often spare the parotids and sublingual glands even when the submandibular glands are
resected as part of a neck dissection. This may minimise the problems with dry mouth which many other
head and neck patients (most notably oropharyngeal and nasopharyngeal cancers) often experience. Several
other studies have also reported that swallowing of solid foods present the greatest challenge and burden to
this group of patients. 19-22. In this study, we excluded patients with recurrent disease and other known
neurological conditions which may perhaps be the group more likely to have problems with swallowing
liquids. When used in clinical practice with a full range of laryngectomy patients (healthy and those with
further/comorbid disease), it is possible that scores on these three items of the SOAL may be worse. It is also
possible that late effects of radiotherapy may result in increased fibrotic tissue and stenosis that in turn may
impede swallowing many years post treatment. 23 These floor effects could therefore shift and until further
work is done maintaining these items is important.
In addition to demonstrating good internal consistency, principal component analysis indicated that all scale
items loaded onto the first component. This confirmed that the same underlying construct is being measured
and that item scores may be combined to give a single score. Furthermore, factor analysis did not support a
8
robust subdomain culture. We also confirmed our hypothesis that individuals with a simple laryngectomy
would have better scores than those with laryngectomy and radiotherapy, and laryngectomy and
chemoradiation. Likewise, as hypothesized patients on a normal textured diet reported better scores than
those on a modified diet. We may therefore speculate that the best SOAL scores will be obtained for those
patients with a simple laryngectomy who are reporting a normal diet. Further work will be necessary before
score boundaries that represent normal post laryngectomy swallow function or mild, moderate, severe
swallowing problems can be determined.
Each item of the SOAL scale includes a rating for ‘bother’ [yes/no] as well as ratings for severity [‘no’ to ‘a
lot’]. The ‘bother ‘ rating was included, as it was evident from the patient focus groups that patients
sometimes experience a symptom but learn to adjust to it and are no longer bothered by it. Other symptoms
may be more bothersome, and they may be more inclined to want help with reducing or minimising these
symptoms. We did not focus on the ‘bother’ ratings in the current analyses, and the role of perceived bother
and patient adjustment will need further investigation. It will be useful to explore the potential contribution
of bother in clinical decision-making, and choosing when to provide intervention.
The high response rate in this study (96.5 %) is reflective of the data collection strategy used. Despite the
lengthy time taken to collect data, this method was chosen in favour of postal surveys that have a poorer
response rate. 24 Further to this, clinicians were required to gather information about treatment modality,
surgical and disease variables from the patient’s medical notes during the clinic visit.
Study limitations and directions for future work
This study was designed so that data could be collected during routine clinical practice and with minimal
disruption to patients and SLT clinicians. In limiting the response burden for patients and the number of
measures collected by clinicians, we omitted to include convergent validity (comparing the SOAL with other
similar validated dysphagia measures) and discriminant validity (comparing the SOAL with different
constructs). We also have a small sample for the test – retest study, and a modest but acceptable sample
size for the factor analysis. However, it has been demonstrated that reliable results can be obtained with
smaller samples. 25 Furthermore, as suggested in the COSMIN checklist 26, it will also be necessary to
establish the measure’s sensitivity and responsiveness to change. A future study with greater resources will
be helpful in more fully addressing these issues.
Despite every effort to ensure a full complement of data, 17 patients were excluded from the known groups
analyses due to incomplete data collection. This was mainly due to incomplete entries in the medical notes
and / or failure to obtain the medical notes during the patients’ clinic visits.
9
We did not systematically collect information on how many patients returned their questionnaires by post
primarily because we had set out to ask all patients to complete the questionnaire in the waiting room. As it
is possible for responses to be affected by different conditions, it will be useful to plan for this contingency in
future work.
Another possible limitation within this study is the absence of data on comorbidities of the patients. We
excluded those patients with neurological conditions associated with possible dysphagia. It may be helpful in
future studies to systematically collect all information that may impact swallowing function. Other descriptive
variables such as social history (alcohol, smoking, marital status) and education level could also be useful.
Finally work is already underway by our own group to examine the swallowing outcomes following total
laryngectomy as measured by the SOAL in relation to the effect of treatment and surgical variables.
Conclusions:
This study has tested the psychometric properties of the SOAL scale in a representative sample of patients
following total laryngectomy and provided strong evidence on its internal consistency and validity. The SOAL
can be used as a research tool to capture information about swallowing function in the laryngectomy
population particularly those under long-term follow-up care. It is easily administered to patients whilst
waiting for routine check-ups in oncology clinics. It has the potential to signpost clinicians to specific areas of
concern regarding a patient’s swallow function, an aspect all too commonly missed in this group of patients
for whom voice restoration is generally the main focus. As is common with new measures, further research
can confirm its psychometric properties and determine its appropriateness as a clinical outcome measure.
Acknowledgement
This study received no specific funding but was made possible by the collaborative efforts of several speech
and language therapists affiliated with the Head & Neck Special Interest Group (South of England). The
authors would like to thank all the clinicians at the participating hospitals who contributed to this project. A
special thanks to Sarah Pilsworth, Lauren Murphy and Carlene Perris for their substantial contribution to the
data collection process.
REFERENCES
10
1. Valderas JM, Kotzeva A, Espallargues M, Guyatt G, et al. The impact of measuring patient-reported
outcomes in clinical practice: a systematic review of the literature. Qual Life Res. 2008;17(2):179–93.
2. Luckett TÃ, Butow PN, King MT. Improving patient outcomes through the routine use of patient-reported
data in cancer clinics : future directions. 2009; 1138:1129–38.
3. Metcalfe CW, Lowe D, Rogers SN. What patients consider important: temporal variations by early and
late stage oral, oropharyngeal and laryngeal subsites. J Craniomaxillofac Surg. 2014;42(5):641–7.
4. Govender R, Breeson L, Tuomainen J, Smith CH. Speech and swallowing rehabilitation following head and
neck cancer : Are we hearing the patient's voice ? Our experience with ten patients. Clin Otolaryngol.
2013; 38:433–7.
5. Rattray J & Jones MC. Essential Elements of questionnaire design. J Clin Nurs. 2007; 16:234-243
6. Sushil S & Verma N. Questionnaire Validation Made Easy. Eur J Sci Res. 2010;46:172–8.
7. Chen AY, Frankowski R, Bishop-Leone J. The development and validation of a dysphagia-specific quality of
life questionnaire for patients with head and neck cancer: The MD Anderson Dysphagia Inventory. Arch.
Otolaryngol.Head Neck Surg. 2001; 127: 870-6.
8. Wallace KL, Middleton S, Cook IJ. Development and validation of a self-report symptom inventory to
assess the severity of oral-pharyngeal gysphagia. Gastroent:2000; 118: 678-87.
9. Dwivedi RC, St Rose S, Roe JWG, Khan As, Pepper C, Nutting CM et al. Validation of the Sydney Swalllow
Questionnaire (SSQ) in a cohort of head and neck cancer patients. Oral Oncol. 2010; 46(4):e10-14
10. McHorney CA, Martin-Harris B, Robbins J, Rosenbeck J. Clinical validity of the SWAL-QOL and SWAL-CARE
outcome tools with respect to bolus flow measures. Dysphagia. 2006; 21(3):141-8
11. Govender R, Lee MT, Davies TC, et al. Development and preliminary validation of a patient-reported
outcome measure for swallowing after total laryngectomy (SOAL questionnaire). Clin Otolaryngol.
2012;37(6):452–9.
12. Ackerstaff AH, Hilgers FJ, Aaronson NK, Balm A J. Communication, functional disorders and lifestyle
changes after total laryngectomy. Clin Otolaryngol Allied Sci. 1994;19(4):295–300.
13. Streiner DL, Norman GR. Health Measurement Scales. A Practical Guide to Their Development and Use.
Oxford: Oxford University Press;2008.
11
14. Lohr KN. Assessing health status and quality-of-life instruments: Attributes and review criteria. Quality of
Life Research. 2002,11:193–205.
15. Lamping DL, Schroter S, Marquis P, Marrel A, Duprat-Lomon I, Sagnier P-P. The community-acquired
pneumonia symptom questionnaire: a new, patient-based outcome measure to evaluate symptoms in
patients with community-acquired pneumonia. Chest. 2002;122(3):920–9.
16. Hilari K, Lamping DL, Smith SC, Northcott S, Lamb A, Marshall J. Psychometric properties of the Stroke
and Aphasia Quality of Life Scale (SAQOL-39) in a generic stroke population. Clin Rehabil. 2009; 23
(6):544–57.
17. Ferguson E, Cox T. Exploratory factor analysis: a user guide. Int J Selection Assess.1993; 1: 84-94.
18. Tabachnick BG, Fidell LS. Using Multivariate Statistics. 6th ed. Boston, Mass: Allyn and Bacon; 2013.
19. Maclean J, Cotton S, Perry A. Post-laryngectomy: it’s hard to swallow: an Australian study of prevalence
and self-reports of swallowing function after a total laryngectomy. Dysphagia. 2009:24 (2):172–9.
20. De Casso C, Slevin NJ, Homer JJ. The impact of radiotherapy on swallowing and speech in patients who
undergo total laryngectomy. Otolaryngol Head Neck Surg. American Academy of Otolaryngology–Head
and Neck Surgery Foundation. 2008;139(6):792–7.
21. Ward EC, Bishop B, Frisby J, Stevens M. Swallowing outcomes following laryngectomy and
pharyngolaryngectomy. Arch Otolaryngol Head Neck Surg. 2002;128(2):181–6.
22. Sullivan P & Hartig G. Dysphagia after total laryngectomy. Current Opinion in Otolaryngology & Head &
Neck Surgery. 2001; 9(3):139-146..
23. Hutcheson KA, Lewin JS, Barringer DA, et al. Late dysphagia after radiotherapy-based treatment of head
and neck cancer. Cancer. 2012;118(23):5793–9.
24. Nakash RA, Hutton JL, Jørstad-Stein EC, Gates S, Lamb SE. Maximising response to postal questionnaires –
A systematic review of randomised trials in health research. BMC Med Res Methodol. 2006; 6:5.
25. Hobart CJ, Cano SJ, Warner TT, Thompson AJ. What sample sizes for reliability and validity in neurology?
Journal of Neurology. 2012; 259(12): 2681-2694.
12
26. Terwee CB, Mokkink LB, Knol DL, Ostelo RWJG, Bouter LM, de Vet HCW. Rating the methodological
quality in systematic reviews of studies on measurement properties: a scoring system for the COSMIN
checklist. Qual Life Res. 2012; 21(4):651-7.
13
Table 1: Participant characteristics
Variable Number of participants = N (%)
Main sample Test / retest N = 110 N = 15
Gender Male 94 (86%) 11 (73.3%) Female 16 (14%) 4 (23.5%) Age Mean (SD) 66 (9.1) 65.2 (4.04) Range 38-90 57-71 T-stage
T3 22 (20%) 2 (13.3%) T4 43 (39.1%) 12 (80%) Salvage 40 (36.7%) 1 (6.6%) Unknown 5 (4.5%)
Flap reconstruction Primary Closure 80 (72.7%) 12 (80%) Radial forearm free flap 2 (1.8%) Pectoralis Major 8 (.2%) 1 (6.6%) Free Jejunum / Gastric pull-up 10 (9%) 1 (6.6%) Unknown 11 (10%) 1 (6.6%) Resection
Larynx alone 66 (60%) 13 (86.7%) Pharynx 25 (22.7%) 1 (6.6%) Tongue Base 4 (3.7%) Other 5 (4.5%) 1 (6.6%)
Closure technique
Horizontal 37 (33.6%) 6 (40%) T-Closure 8 (7.3%) 3 (20%) Vertical 6 (5.5%) 0 Unknown 50 (45.5%) 4 (26.7%) Not applicable 9 (8.2%) 2 (13.3%)
Layers of closure
Two 31 (28.2%) 2 (13.3%) Three 26 (23.6%) 4 (26.7%) Unknown 53 (48.2%) 9 (60%)
Myotomy
Yes 71 (64.5%) 8 (53.3%) No/Not applicable 14 (12.7%) 2 (13.3%) Unknown 25 (22.7%) 5 (33.3%)
14
Table 1 continued: Participant characteristics
Variable Number of participants = N (%)
Main sample Test / retest N = 110 N = 15
Additional reported interventions Dilatation 20 (18.2%) 1 (6.6%) Botox 10 (9.1%) 0
Time post surgery (in months) Mean [SD] 61 [65.2] 66 [46.8] Median [IQR] 39 [12-84] 60 [22-103] Range 3-252 12-156 Diet Normal 77 (70%) 12 (80%) Modified 31 (28%) 3 (20%) No oral intake 1 (1%) Missing data 2 (2%) Treatment group Laryngectomy only 12 (11%) 4 (26.7%) Laryngectomy and radiotherapy 59 (54%) 6 (40%) Laryngectomy and chemo-radiation 22 (20%) 5 (33.3%)
Missing data 17 (15%)
15
Table 2: Mean(SD) and selected psychometric properties of the Swallowing Outcomes After Laryngectomy (SOAL) measure
Results
Mean (SD)a 11.3 (7.6) Sample score range (possible range) 0 – 34 (0 – 34)
Missing datab (> 10%) 0 items Ceiling effectsb (>80%) 0 items
Floor effectsb (>80%) 3 items (17%) Skewnessb (> ±1) 5 items (29%)
Cronbach’s alphaa 0.91 Item-total correlationsa 0.38 - 0.77
Test-retest reliabilityc Intra-class correlation coefficient 0.73 NOTE: aN=105; bN=110; cN=15
Table 3: Known groups validity of SOAL Group N Mean SD df t-statistic p
Diet Normal diet
74 8.6 6.0 102 -7.33 <0.001
Modified or no oral diet
30 18.3 6.5
Treatment F-ratio Laryngectomy
12 7.4 7.7 2, 85 8.02 0.001
Laryngectomy + radiotherapy
56 10.2 6.2
Laryngectomy + chemoradiation
20 16.6 8.8
Recommended